Commit4753ef3

committed

Use a WaitLatch for vacuum/autovacuum sleeping

Instead of using pg_usleep() in vacuum_delay_point(), use a WaitLatch.This has the advantage that we will realize if the postmaster has beenkilled since the last time we decided to sleep while vacuuming.Reviewed-by: Thomas MunroDiscussion:https://postgr.es/m/CAFh8B=kcdk8k-Y21RfXPu5dX=bgPqJ8TC3p_qxR_ygdBS=JN5w@mail.gmail.com

1 parent54bb91c commit4753ef3Copy full SHA for 4753ef3

File tree

-3

lines changed

-3

lines changed

Lines changed: 5 additions & 3 deletions

Original file line number	Diff line number	Diff line change
`@@ -2080,9 +2080,11 @@ vacuum_delay_point(void)`
`2080`	`2080`	`if (msec>VacuumCostDelay*4)`
`2081`	`2081`	`msec=VacuumCostDelay*4;`
`2082`	`2082`
`2083`		`-pgstat_report_wait_start(WAIT_EVENT_VACUUM_DELAY);`
`2084`		`-pg_usleep((long) (msec*1000));`
`2085`		`-pgstat_report_wait_end();`
	`2083`	`+(void)WaitLatch(MyLatch,`
	`2084`	`+WL_LATCH_SET \|WL_TIMEOUT \|WL_EXIT_ON_PM_DEATH,`
	`2085`	`+msec,`
	`2086`	`+WAIT_EVENT_VACUUM_DELAY);`
	`2087`	`+ResetLatch(MyLatch);`
`2086`	`2088`
`2087`	`2089`	`VacuumCostBalance=0;`
`2088`	`2090`

Comments

(0)